Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in [1].
Adaptive Probing Policies for Shortest Path Routing
Authors: Aditya Bhaskara, Sreenivas Gollapudi, Kostas Kollias, Kamesh Munagala
NeurIPS 2020 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We verify these assumptions and validate the performance of our algorithms on real data. (Abstract) In Section 3, we empirically demonstrate how to overcome these obstacles on real-world data, and reduce the overall probing problem to the parallel edge case. (Section 1.1) |
| Researcher Affiliation | Collaboration | Aditya Bhaskara School of Computing University of Utah EMAIL Sreenivas Gollapudi Google Research Mountain View, CA EMAIL Kostas Kollias Google Research Mountain View, CA EMAIL Kamesh Munagala Department of Computer Science Duke University EMAIL |
| Pseudocode | Yes | Algorithm 1 Greedy probing |
| Open Source Code | No | The paper does not provide any concrete access (link, explicit statement) to the source code for the methodology described. |
| Open Datasets | Yes | This dataset contains hourly average traffic speeds on road segments throughout New York City. It covers four years of traffic estimates in New York City estimated from approximately 700 million taxi trips from 2010-2013 [Donovan and Work, 2017]. |
| Dataset Splits | No | The paper mentions evaluating performance on data, but does not provide specific details on train/validation/test splits of the dataset with percentages or counts for reproducibility. |
| Hardware Specification | No | The paper does not provide specific hardware details (e.g., GPU/CPU models, memory amounts) used for running its experiments. |
| Software Dependencies | No | The paper does not provide specific ancillary software details (e.g., library or solver names with version numbers) needed to replicate the experiment. |
| Experiment Setup | Yes | The setup is as follows. For each source-destination pair, we fix a canonical path p0 and its length (at a certain time). For every other path p, the distribution Xp used is the discrete, empirical distribution of its path length in all past time steps where length of p0 is within 5% of . We use two different values for parameter δ : δ = 0.1 and δ = 0.01. We set = 0, so that we seek an exact shortest path with probability 1 δ . |